Abstract

An unmanned surface vehicle (USV) plans its global path before the mission starts. When dynamic obstacles appear during sailing, the planned global path must be adjusted locally to avoid collision. This study proposes a local path planning algorithm based on the velocity obstacle (VO) method and modified quantum particle swarm optimization (MQPSO) for USV collision avoidance. The collision avoidance model based on VO not only considers the velocity and course of the USV but also handles the variable velocity and course of an obstacle. According to the collision avoidance model, the USV needs to adjust its velocity and course simultaneously to avoid collision. Due to the kinematic constraints of the USV, the velocity window and course window of the USV are determined by the dynamic window approach (DWA). In summary, local path planning is transformed into a multiobjective optimization problem with multiple constraints in a continuous search space. The optimization problem is to obtain the USV’s optimal velocity variation and course variation to avoid collision and minimize its energy consumption under the rules of the International Regulations for Preventing Collisions at Sea (COLREGs) and the kinematic constraints of the USV. Since USV local path planning is completed in a short time, it is essential that the optimization algorithm can quickly obtain the optimal value. MQPSO is primarily proposed to meet that requirement. In MQPSO, the efficiency of quantum encoding in quantum computing and the optimization ability of representing the motion states of the particles with wave functions to cover the whole feasible solution space are combined. Simulation results show that the proposed algorithm can obtain the optimal values of the benchmark functions and effectively plan a collision-free path for a USV.

Highlights

  • An unmanned surface vehicle (USV) is an autonomous marine vehicle that has emerged as a viable tool to perform tasks that are dangerous or unsuitable for manned vessels

  • Is study proposes a local path planning algorithm for USV collision avoidance based on modified quantum particle swarm optimization (MQPSO). e main contributions are as follows: (1) Based on the model proposed in [7, 17], a USV collision avoidance model is established by the velocity obstacle (VO) method considering the constraints of USV kinematic model, the rules of COLREGs, and the uncertainty of obstacle’s velocity. e collision avoidance model considers the velocity and course of the USV and handles the variable velocity and course of an obstacle

  • Due to the kinematic constraints of the USV, the velocity window and course window of the USV are determined by dynamic window approach (DWA). erefore, the local path planning for USV collision avoidance is transformed into a multiobjective optimization problem with multiple constraints in a continuous search space

Read more

Summary

Introduction

An unmanned surface vehicle (USV) is an autonomous marine vehicle that has emerged as a viable tool to perform tasks that are dangerous or unsuitable for manned vessels. Is study proposes a local path planning algorithm for USV collision avoidance based on modified quantum particle swarm optimization (MQPSO). Erefore, the local path planning for USV collision avoidance is transformed into a multiobjective optimization problem with multiple constraints in a continuous search space. To avoid premature convergence and local optima, a new optimization algorithm called modified quantum particle swarm optimization (MQPSO), combining QBPSO and quantum computing, is proposed in this study.

Problem Statement
Collision Avoidance Model
Optimization Algorithm
Evaluation and Simulation Studies
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call